The Maintenance of Key Pecking by Stimulus‐contingent and Response‐independent Food Presentation

Abstract

Three naive pigeons were exposed to a series of two-component multiple schedules of response-independent food presentation. The component schedules were sometimes identical (non-differential procedures) and sometimes different (differential procedures). High rates of key pecking were nmaintained in all the differential procedures, and pecking decreased substantially in non-differential procedures, even when the frequency of food presentation in non-differential procedures was higher than in differential procedures. It is suggested that the high rates of key pecking were maintained not by adventitious response-reinforcer contingencies, but by differential contingencies between the stimulus (keylight) and food.The role of such contingencies in the phenomenon of behavioral contrast is discussed.If a pigeon that has been trained to peck on an illuminated key to produce food is then exposed to a procedure in which food delivery occurs witlhout regard to its behavior (response-independent food presentation), the response is nevertheless maintained at nonzero rates (Appel and Hiss, 1962;Edwards, West, and Jackson, 1968;Herrnstein, 1966;Herrnstein and Morse, 1957;Lachter, 1971;Neuringer, 1970;Zeiler, 1968). Similar findings have been reported with rats and bar pressing (Edwards, et al., 1968;Lattal and Mazey, 1971;Rescorla and Skucy, 1969; Skinner, 1938, pp. 163-166). This phenomenon has typically been explained as a corollary of the effects of response-dependent reinforcement (e.g., Herrnstein, 1966 again precede it. In this fashion, responseindependent food presentation can maintain a given class of behavior. However, since an explicit dependency between the behavior and the food is absent, the organism at times will obtain food with an appreciable delay since the last occurrence of the behavior in question, or immediately after engaging in any of a number of different behaviors. Thus, behavior other than the one being measured will also be strengthened, which is consistent with the observation that responding is maintained at a lower rate by response-independent than by response-dependent food presentation.However, an alternative account of the effectiveness of response-independent food presentation in maintaining key pecking in the pigeon has been offered by Staddon and Simmellhag (1971). They argued that the presence of food in a situation, combined with a pigeon's state of food deprivation, will induce the pigeon to peck in that situation. Moreover, they reported that while pigeons engaged in a wide variety of behaviors in the time intervals between response-independent food presentations, in the time just before delivery of the food, all pigeons were pecking. They further suggested that the rate of pecking is no lower when food delivery is responseindependent than when it is response-dependent: the difference in pecking between the two situations is that the locus of the peck is more variable when food delivery is response-1973, 19, 65-72 NUMBER I (JANUARY)

Multiple-schedule interactions and discrimination

“…Finally, even in cases in which behavioral contrast did occur in S-/S-components, the absolute response rates associated with these components may have been so low that relatively large increases went unnoticed. Whatever the case, the observation that S-/S-responding may show positive behavioral contrast has implications for several interpretations of behavioral contrast, such as "relative reinforcement theory" (Reynolds, 1961b), "additivity theory" (Gamzu & Schwartz, 1973;Schwartz & Gamzu, 1977),and "behavioral competition theory" (Hinson & Staddon, 1978). Reynolds (1961 b) suggested that positive behavioral contrast results from an increase in the relative reinforcement frequency associated with the stimulus showing contrast.…”

“…Additivity theory (Gamzu & Schwartz, 1973;Schwartz & Gamzu, 1977) also has troubles accounting for positive behavioral contrast in S-/S-responding. According to additivity theory, autopecks are elicited by the presentation of a stimulus when that stimulus is "differentially associated with a higher frequency of food presentations than obtains in the absence of that stimulus" (Gamzu & Schwartz, 1973, p. 70).…”

Simultaneous discrimination learning: Stimulus interactions

“…Finally, Gamzu and Schwartz (1973) suggested that response rate to a stimulus is determined by two additive factors-first, the operant contingency of reinforcement associated with responding and, second, the Pavlovian relation of the stimulus to reinforcement. According to Gamzu and Schwartz, it is the Pavlovian relation that is affected by the introduction of the poorer schedule and that produces contrast.…”

Responding on a multiple variable-time variable-time schedule as a function of component duration

“…They pecked at a lower rate during the stimulus that signaled the lower rate of food (the less favorable stimulus). And they pecked both stimuli at a lower rate when the stimuli signaled equal rates of food presentation.Several recent, additive, theories have related these responses to the behavioral contrast which occurs when pigeons respond on multiple variable-interval variable-interval (mult VI VI) schedules of reinforcement (e.g., Gamzu & Schwartz, 1973; Hearst & Jenkins, 1974;Rachlin, 1973). Mult VI VI schedules are similar to mult VT VT schedules, except that food is not given freely.…”

“…Several recent, additive, theories have related these responses to the behavioral contrast which occurs when pigeons respond on multiple variable-interval variable-interval (mult VI VI) schedules of reinforcement (e.g., Gamzu & Schwartz, 1973; Hearst & Jenkins, 1974;Rachlin, 1973). Mult VI VI schedules are similar to mult VT VT schedules, except that food is not given freely.…”